变化条件下的黄河上游梯级水库冰期调度优化

The reservoir system in the upper Yellow River plays a vital role in controlling ice floods for the Ningxia-Inner Mongolia Reach. In recent years, the circumstance of Ningxia-Inner Mongolia Reach changes dramatically: the extreme climate events, severe river channel sedimentation as well as complicated ice characteristics have restricted efficient utilization of water and hydropower resources. In this study, through analyzing historical hydrological and meteorological data of multiple years, the evolution rules of ice condition and its response mechanism to climate change and human activities will be identified; the critical reservoir operation rules will be explored at different stages of an ice period (i.e. ice drift stage, freeze stage, breakup stage); the characteristics of sediment transport during ice period will be described; and further the feasibility for increasing river channel scour of Inner Mongolia Reach by ice flood peak as well as reservoir system operation during breakup stage will be initially demonstrated. An optimization model that ensures ice flood control and considers hydropower generation, water resources utilization and sediment transport efficiency for the reservoir system in the upper Yellow River will be formulated. Accordingly, a parallel optimization method will be developed under a high-performance computing environment. The computations will be run for the present and future situations under different combinations of climate, bankfull discharge and inflow frequency. Furthermore, the operation rule of multiple years for the Longyangxia Reservoir, the reserve storage capacity for the Liujiaxia Reservoir, and the optimal operation rule for the reservoir system during ice period will be determined. The study will provide technical supports for improving the joint operation of the reservoir system in the upper Yellow River during ice period, and further, scientifically addressing climate change, fully avoiding potential ice disasters, efficiently utilizing water and hydropower resources as well as improving water-sediment conditions.

黄河上游梯级水库承担着宁夏—内蒙古河段（“宁蒙河段”）冰期减灾任务。近年来宁蒙河段环境不断变化，极端气候频现、河床严重淤积、冰凌机理复杂等制约了水和水能资源高效利用。本研究通过分析历史多年水文气象数据，识别宁蒙河段河冰对气候变化和人类活动的响应机制，挖掘流凌、封河、开河等冰期不同阶段水库临界运用方式；揭示和描述宁蒙河段冰期输沙规律及特征，初步论证利用开河期凌峰与梯级水库调度对接实现内蒙古河段塑槽的可行性；通过建立确保防凌、兼顾发电、用水、输沙等目标的黄河上游梯级水库冰期调度模型，开发与之相适的并行优化求解算法，计算现状和预测未来不同气候条件、平滩流量、来水频率组合情景下龙羊峡水库多年运用方式、刘家峡水库预留防凌库容、冰期梯级水库联合运用方案等。本研究成果将为改进黄河上游梯级水库冰期运用方式，从而科学应对气候变化、充分发挥防凌减灾效益、高效利用水和水能资源、改善水沙条件提供科技支撑。

黄河上游梯级水库承担着宁夏—内蒙古河段（“宁蒙河段”）防洪防凌、西北电网电力电量平衡、保障宁蒙灌区农业灌溉及黄河流域中下游地区供水、生态、输沙等综合运用任务。近年来，全球变暖和河流梯级开发造成气象水文特征剧烈变化，沿河社会经济发展用水需求攀升与水资源短缺矛盾愈加深化，为上游梯级水电站调度管理带来了极为严峻的挑战。为此，本项目开展研究工作并取得主要成果如下：（1）收集整理了宁蒙河段的历史水文气象数据，研究了河冰特征与气温的相关关系，宁蒙河段气温以2.19℃/50年速率升高，最早流凌时间以8.1天/50年速率延迟，最早封河时间以4.9天/50年速率延迟，最早开河时间以26.9天/50年速率提前，导致封河历时以7.1天/50年速率缩短，冰厚以0.6厘米/年速率变薄，河冰特征和气温的变点均发生在1980年代中期；（2）收集整理了黄河上游地区水资源公报、已建水库特征参数和历史调度资料，在了解不同地区、不同行业取用水情况和主要水库运行方式基础上，定量描述了防洪防凌、水力发电、农业灌溉、生态保障、输沙水量、沿河供水等水库调度控制目标和约束，构建了梯级水库多目标协同调度模型，提出了基于LINGO的水库优化调度决策支持框架；（3）基于高性能计算平台，研发了水库群调度的并行优化求解算法，包括并行异步混合算法和考虑计算资源经济可行的并行动态规划方法，能够混合不同算法的优势特征，利用分布式存储和分布式计算，解决大规模、高维性、动态性、非线性等难题，兼顾求解效率和求解质量；（4）研究了不同情景、不同边界条件下的梯级水库联合运行方式，揭示了接入青海电网的黄河水电极大性与稳定性的权衡关系，黄河上游流域水能粮纽带效益关系，水库防汛发电输沙三目标互馈关系，为改进梯级水库运用方式，从而科学应对变化环境、充分发挥梯级水库多目标运行效益、高效利用水和水能资源等提供科技支撑。相关研究成果出版学术专著1部，发表SCI/EI期刊论文11篇，国际会议论文1篇，授权国家发明专利2项。